The use of highly sophisticated integrated hydrologic and hydraulic modeling tools for the design of stormwater facilities in Municipal Separate Storm Sewer System (MS4) communities has resulted in profound improvements in
the performance of these facilities. However, the stormwater facilities in an MS4 are rarely isolated and often are a part of a complex hydraulic network consisting of pipes, storage facilities, pumps and other elements. Engineers have traditionally employed a brute-force approach to identify the most optimal design solution for a stormwater problem by iteratively running a limited set of alternative options. The traditional approach of selectively running alternatives by a manual approach is extremely time consuming for a complex system and the results achieved are still not optimal. For a given system, thousands of iterations are required; however, budget and time constraints usually limit the analysis of less than a hundred options.
The use of a Genetic Algorithm Optimization technique is an evolving approach to truly design and identify an optimal solution for a complex MS4 stormwater system. The algorithm starts with a sample set of solutions and develops a new generation of solutions by eliminating the non-optimal solutions and combining traits of near optimal solutions. This process is repeated over and over until the suite of optimal solutions are found. With this technique, hundreds of thousands of alternatives can be evaluated and compared. The presented case study included over 200,000 alternatives.
The presented project involved evaluating a multi-faceted drainage system that collects approximately 356 acres of urban runoff in an MS4 community. The system is comprised of a network of streams, ponds, detention basins, and major trunk sewers. A watershed approach was taken for this project to effectively solve flooding issues and improve water quality to support an intelligent capital improvement plan (CIP) to support MS4 compliance. The use of a genetic algorithm optimization technique allowed the community to develop a multi-year phased CIP that achieved optimal performance by prioritizing improvements, while meeting annual budgetary constraints with an adaptive long term implementation strategy.